Dystocia refers to slow progress in the first stage of labour. The incidence in nulliparae is 20% and if untreated can lead to maternal and neonatal morbidity. Augmentation of labour to prevent and treat dystocia includes amniotomy and intravenous syntocinon. However, the optimal timing of such intervention is controversial, and leaves little scope for maternal choice, which underpins current maternity care. The literature on the physiology of labour indicates that it is a complex process, which is not fully understood. The onset and progress of labour is defined in terms of precise cervical dilations, however it is likely that there is a range of dilations consistent with 'normal' labour. The current definition of dystocia may be too restrictive and a management option, which delays obstetric intervention, might facilitate greater opportunity for normal progress. The hormonal pathways that regulate labour have some commonality with those of the stress response. This thesis presents a study designed to examine whether labouring in water, a management option that may alleviate some of the stress of labour, in particular that due to pain, might also facilitate labour progress, and affect the key labour outcomes of epidural analgesia and operative delivery. The limited literature on labouring in water suggests it is an appropriate care option for nulliparae who are at low risk of complications. Labouring in water has been associated with a reduction in pharmacological analgesia, however there were no studies involving women with dystocia. Prior to initiating a full randomised controlled trial to test the effect of labouring in water on labour outcomes for women with dystocia, a two part feasibility study was undertaken. A pilot randomised controlled trial indicated that it was feasible to conduct the planned study in the clinical area but with two trial arms (water and augmentation), instead of the proposed three arms, (water, conservative, and augmentation). A case audit formed the second part of the feasibility study. Data were used to estimate the incidence of dystocia, likely accrual rates, the potential magnitude of any change in epidural analgesia and operative delivery rates, and the required sample size for the main study. The main RCT was conducted during 1999 and 2000. A case audit over the same timeframe provided comparative data and monitored concurrent changes in practice. The epidural analgesia rate was 47% in women allocated to labour in water compared to 66% in women in the augmentation arm (OR 0.46, 95%CI 0.20 to 1.03. P=0.56). There was no difference in the operative delivery rates for women allocated to labour in water compared to those who were not. There was a significant reduction in obstetric interventions (OR 0.10 , 95% CI 0.22 to 0.49, P=0.000). The main limitation of the trial was a lower than expected recruitment rate, with 99 women being randomised compared to the intended 220. This was partially due to changes in practice during the second year of the trial. Despite the small sample, the implications for practice include the need for further consideration of more flexible definitions for labour onset and progress, and the use of labour in water for women as a supportive management option where a delay in obstetric intervention is advocated.